文献速递皮肤肿瘤

真实临床场景下人工智能皮肤癌诊断的局限性

Limits of Artificial Intelligence Models for Skin Cancer Diagnosis in Realistic Settings.
PubMed 文献 · JAMA Dermatol · 2026-01-01

该多机构诊断研究比较了三种AI算法与不同经验皮肤科医生对1117例皮肤病变图像的诊断准确性,发现经验>10年专家多分类准确率最高(74.2%),优于所有AI模型;单模态基础模型仅超过经验<3年读者,与3-10年经验者相当。

要点

  • 所有人类读者平均准确率65.9%,高于CNN的56.7%。
  • 单模态模型准确率72.2%,超过经验<3年读者(68.2%)。
  • 经验>10年专家准确率74.2%,优于所有AI模型。

核心结论

  • 现代基础模型在皮肤病变诊断准确率上超过经验<3年的读者,与3-10年经验者相当,但仍低于经验>10年的专家。

研究设计与样本

  • 多机构诊断研究,使用包含1117例临床和皮肤镜图像及元数据的数据集,纳入时间2023年3月16日至2025年8月1日。
  • 三种AI算法:第一代卷积神经网络(CNN)、PanDerm单模态和多模态基础模型;人类读者评估从同一数据集中抽取的100例分层随机病例。
  • 共652名医生(中位年龄33岁,85.7%女性)贡献了1092次测试迭代。

主要发现

  • 经验>10年专家平均准确率74.2%(5.7%),多模态模型为66.3%(3.8%),专家在多分类诊断主要终点上优于所有AI模型。

临床怎么用

  • AI模型在真实临床场景中诊断皮肤病变时,对经验<3年者可能有辅助价值,但尚不能替代经验>10年的专家。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

IMPORTANCE: Artificial intelligence (AI) systems for skin cancer detection perform well in controlled settings but frequently underperform in everyday clinical practice, raising critical questions about their readiness for deployment. OBJECTIVE: To compare the diagnostic accuracy of AI algorithms vs human evaluators across varying expertise levels for skin lesion diagnosis, including rare and atypical cases, in a realistic clinical context. DESIGN, SETTING, AND PARTICIPANTS: This multi-institutional diagnostic study compared diagnostic performance among AI models and physician readers with varying dermatological expertise, ranging from less than 1 year to more than 10 years of experience. A dataset of dermatological images representing everyday clinical scenarios was used and contained 1117 cases, including clinical and dermoscopic images with associated metadata. Study inclusion spanned from March 16, 2023, to August 1, 2025. EXPOSURES: Three AI algorithms: a first-generation convolutional neural network (CNN) and 2 foundation models (PanDerm unimodal and multimodal). Human readers evaluated 100 stratified, random cases drawn from the same dataset. MAIN OUTCOMES AND MEASURES: The primary outcome was reader-level multiclass diagnostic accuracy for skin lesion classification. Secondary outcomes were binary benign vs malignant sensitivity, specificity, and balanced accuracy. Performance was compared between AI algorithms and human readers stratified by experience level. RESULTS: A total of 652 physicians (median [IQR] age, 33 [29-37] years; 559 [85.7%] female) contributed to 1092 test iterations. All human readers outperformed the CNN (mean [SD] accuracy, 65.9% [10.5%] vs 56.7% [3.9%]; difference, 9.2 percentage points [pp]; 95% CI, -9.8 to 8.5 pp; P < .001). Unimodal accuracy exceeded readers with less than 3 years of experience (mean [SD] accuracy, 72.2% [3.5%] vs 68.2% [7.6%]; difference, 4.0 pp; 95% CI, 3.2-4.9 pp; P < .001). With a mean (SD) accuracy of 74.2% (5.7%), experts with more than 10 years of experience achieved the highest multiclass diagnostic accuracy, outperforming all AI models on this primary end point, which included 56.7% (3.9%) for CNN, 72.2% (3.5%) for the unimodal model, and 66.3% (3.8%) for the multimodal model. CONCLUSIONS AND RELEVANCE: In this diagnostic study, a modern foundation model surpassed readers with less than 3 years of experience on accuracy of skin lesion diagnosis and matched those with 3 to 10 years of experience but remained inferior to experts with more than 10 years of experience, highlighting both the promise and current limitations of AI in dermatologic diagnosis.

来源:JAMA Dermatol,查看原文。版权归原作者所有。

引用

  1. Anriot J, Yan S, Coste C, et al. Limits of Artificial Intelligence Models for Skin Cancer Diagnosis in Realistic Settings. JAMA Dermatol. 2026. PMID: 42234423. doi:10.1001/jamadermatol.2026.1492

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。